Accelerate GPU Concurrent Kernel Execution by Mitigating Memory Pipeline Stalls.
Hongwen DaiZhen LinChao LiChen ZhaoFei WangNanning ZhengHuiyang ZhouPublished in: HPCA (2018)
Keyphrases
- concurrent execution
- distributed shared memory
- multithreading
- memory management
- multi threaded
- real time
- kernel function
- kernel methods
- resource consumption
- support vector
- mutual exclusion
- graphics processors
- memory usage
- memory requirements
- graphics hardware
- intel xeon
- parallel computation
- feature space
- parallel processing
- data transfer
- concurrent processes
- main memory
- memory size
- computational power
- parallel computing
- data flow
- message passing
- parallel implementation
- efficient implementation
- gpu accelerated
- memory bandwidth
- parallel hardware
- parallel programming
- dynamically created
- petri net
- graphics processing units